LU503042B1 - Multi-model learning particle swarm-based intelligent city signal light timing optimization method - Google Patents
Multi-model learning particle swarm-based intelligent city signal light timing optimization method Download PDFInfo
- Publication number
- LU503042B1 LU503042B1 LU503042A LU503042A LU503042B1 LU 503042 B1 LU503042 B1 LU 503042B1 LU 503042 A LU503042 A LU 503042A LU 503042 A LU503042 A LU 503042A LU 503042 B1 LU503042 B1 LU 503042B1
- Authority
- LU
- Luxembourg
- Prior art keywords
- particle
- exemplar
- dimension
- traffic signal
- entrance
- Prior art date
Links
- 239000002245 particle Substances 0.000 title claims abstract description 100
- 238000005457 optimization Methods 0.000 title claims abstract description 34
- 238000000034 method Methods 0.000 title claims abstract description 33
- 238000012937 correction Methods 0.000 claims description 3
- 238000011156 evaluation Methods 0.000 claims description 3
- 230000002708 enhancing effect Effects 0.000 description 2
- 101100366941 Homo sapiens STON1 gene Proteins 0.000 description 1
- 102100021683 Stonin-1 Human genes 0.000 description 1
- 230000004075 alteration Effects 0.000 description 1
- 230000009286 beneficial effect Effects 0.000 description 1
- 238000011161 development Methods 0.000 description 1
- 238000010801 machine learning Methods 0.000 description 1
- 238000012986 modification Methods 0.000 description 1
- 230000004048 modification Effects 0.000 description 1
- 238000005070 sampling Methods 0.000 description 1
- 238000006467 substitution reaction Methods 0.000 description 1
Classifications
-
- G—PHYSICS
- G08—SIGNALLING
- G08G—TRAFFIC CONTROL SYSTEMS
- G08G1/00—Traffic control systems for road vehicles
- G08G1/01—Detecting movement of traffic to be counted or controlled
- G08G1/0104—Measuring and analyzing of parameters relative to traffic conditions
- G08G1/0137—Measuring and analyzing of parameters relative to traffic conditions for specific applications
- G08G1/0145—Measuring and analyzing of parameters relative to traffic conditions for specific applications for active traffic flow control
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
- G06N3/00—Computing arrangements based on biological models
- G06N3/004—Artificial life, i.e. computing arrangements simulating life
- G06N3/006—Artificial life, i.e. computing arrangements simulating life based on simulated virtual individual or collective life forms, e.g. social simulations or particle swarm optimisation [PSO]
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
- G06N3/00—Computing arrangements based on biological models
- G06N3/12—Computing arrangements based on biological models using genetic models
- G06N3/126—Evolutionary algorithms, e.g. genetic algorithms or genetic programming
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06Q—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES; SYSTEMS OR METHODS SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES, NOT OTHERWISE PROVIDED FOR
- G06Q10/00—Administration; Management
- G06Q10/04—Forecasting or optimisation specially adapted for administrative or management purposes, e.g. linear programming or "cutting stock problem"
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06Q—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES; SYSTEMS OR METHODS SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES, NOT OTHERWISE PROVIDED FOR
- G06Q50/00—Information and communication technology [ICT] specially adapted for implementation of business processes of specific business sectors, e.g. utilities or tourism
- G06Q50/10—Services
- G06Q50/26—Government or public services
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06Q—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES; SYSTEMS OR METHODS SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES, NOT OTHERWISE PROVIDED FOR
- G06Q50/00—Information and communication technology [ICT] specially adapted for implementation of business processes of specific business sectors, e.g. utilities or tourism
- G06Q50/40—Business processes related to the transportation industry
-
- G—PHYSICS
- G08—SIGNALLING
- G08G—TRAFFIC CONTROL SYSTEMS
- G08G1/00—Traffic control systems for road vehicles
- G08G1/07—Controlling traffic signals
- G08G1/08—Controlling traffic signals according to detected number or speed of vehicles
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
- G06N20/00—Machine learning
-
- G—PHYSICS
- G08—SIGNALLING
- G08G—TRAFFIC CONTROL SYSTEMS
- G08G1/00—Traffic control systems for road vehicles
- G08G1/01—Detecting movement of traffic to be counted or controlled
- G08G1/0104—Measuring and analyzing of parameters relative to traffic conditions
- G08G1/0108—Measuring and analyzing of parameters relative to traffic conditions based on the source of data
- G08G1/0116—Measuring and analyzing of parameters relative to traffic conditions based on the source of data from roadside infrastructure, e.g. beacons
-
- G—PHYSICS
- G08—SIGNALLING
- G08G—TRAFFIC CONTROL SYSTEMS
- G08G1/00—Traffic control systems for road vehicles
- G08G1/01—Detecting movement of traffic to be counted or controlled
- G08G1/0104—Measuring and analyzing of parameters relative to traffic conditions
- G08G1/0125—Traffic data processing
- G08G1/0129—Traffic data processing for creating historical data or processing based on historical data
-
- Y—GENERAL TAGGING OF NEW TECHNOLOGICAL DEVELOPMENTS; GENERAL TAGGING OF CROSS-SECTIONAL TECHNOLOGIES SPANNING OVER SEVERAL SECTIONS OF THE IPC; TECHNICAL SUBJECTS COVERED BY FORMER USPC CROSS-REFERENCE ART COLLECTIONS [XRACs] AND DIGESTS
- Y02—TECHNOLOGIES OR APPLICATIONS FOR MITIGATION OR ADAPTATION AGAINST CLIMATE CHANGE
- Y02A—TECHNOLOGIES FOR ADAPTATION TO CLIMATE CHANGE
- Y02A30/00—Adapting or protecting infrastructure or their operation
- Y02A30/60—Planning or developing urban green infrastructure
Landscapes
- Engineering & Computer Science (AREA)
- Business, Economics & Management (AREA)
- Physics & Mathematics (AREA)
- General Physics & Mathematics (AREA)
- Theoretical Computer Science (AREA)
- Health & Medical Sciences (AREA)
- Tourism & Hospitality (AREA)
- Economics (AREA)
- Human Resources & Organizations (AREA)
- Strategic Management (AREA)
- General Health & Medical Sciences (AREA)
- Marketing (AREA)
- Biophysics (AREA)
- General Business, Economics & Management (AREA)
- Life Sciences & Earth Sciences (AREA)
- Primary Health Care (AREA)
- Development Economics (AREA)
- Educational Administration (AREA)
- Mathematical Physics (AREA)
- Biomedical Technology (AREA)
- Evolutionary Biology (AREA)
- Computational Linguistics (AREA)
- Data Mining & Analysis (AREA)
- Evolutionary Computation (AREA)
- Bioinformatics & Computational Biology (AREA)
- Molecular Biology (AREA)
- Computing Systems (AREA)
- General Engineering & Computer Science (AREA)
- Artificial Intelligence (AREA)
- Software Systems (AREA)
- Bioinformatics & Cheminformatics (AREA)
- Game Theory and Decision Science (AREA)
- Chemical & Material Sciences (AREA)
- Analytical Chemistry (AREA)
- Entrepreneurship & Innovation (AREA)
- Operations Research (AREA)
- Quality & Reliability (AREA)
- Physiology (AREA)
- Genetics & Genomics (AREA)
- Traffic Control Systems (AREA)
Abstract
The present invention discloses a method for optimizing smart city traffic signal timing based on multi-exemplar learning particle swarms, and mainly relates to the fields of smart traffic control and intelligent optimization algorithms. Aiming at the problem that the traditional particle swarm optimization algorithm is prone to fall into the local optimum when applied to the traffic signal timing optimization scene, this method adopts a new multi-exemplar learning strategy, which enables particles to learn from exemplars of other particles in different dimensions while learning from their own optimal position and global optimal position, helping to enhance the algorithm diversity and avoid falling into the local optimal position. In addition, this method adopts a knowledge embedding auxiliary strategy when generating the initial particle population, and uses the distribution characteristics of the input traffic flow at the intersection as the prior-knowledge to assist the generation of the initial population. The optimization results of the traffic flow at a single intersection with different saturation levels show that this method, compared with other timing optimization methods, has better diversity on the premise of ensuring the convergence velocity, and can obtain an optimized timing scheme with better comprehensive performance.
Description
MULTI-MODEL LEARNING PARTICLE SWARM-BASED INTELLIGENT CITY/508042
SIGNAL LIGHT TIMING OPTIMIZATION METHOD
The present invention relates to the fields of smart traffic control and intelligent optimization algorithms, specifically to a method for optimizing smart city traffic signal timing based on multi-exemplar learning particle swarms.
With the development of urbanization, the number of cars owned by urban residents is increasing year by year, and the traffic pressure faced by urban roads, especially intersections, is also gradually increasing. Setting traffic signals can strengthen road traffic management and effectively solve the problem of traffic flow conflicts, but with the increase of traffic flow saturation, the road traffic situation becomes increasingly complex. The traditional traffic signal timing scheme, such as the Webster timing scheme, establishes a traffic mathematical exemplar based on the goal of minimizing the intersection delay, then calculates the optimal traffic signal cycle based on this exemplar, and then divides the corresponding green signal time. However, the Webster model is only applicable to the traffic exemplar for the unsaturated traffic flow. Therefore, when the road traffic flow is oversaturated, the Webster timing scheme is no longer applicable.
The traditional traffic signal timing scheme has become less and less able to meet the needs of smart city traffic.
As an important evolutionary optimization algorithm, the particle swarm optimization algorithm has shown excellent performance in various fields such as power system, medical image registration, multi-objective optimization, and machine learning. The particle swarm optimization algorithm has strong global search ability and convergence ability, and is suitable for the search optimization of the best traffic signal timing scheme under different saturation conditions.
Therefore, many researchers also apply the particle swarm optimizatfor?3042 algorithm to the field of smart traffic signal timing optimization.
Most of the existing particle swarm optimization algorithms for traffic signal optimization have the shortcoming that they are prone to fall into the local optimal solution. This is because the particle swarm searches for the optimal value in a continuous space, while the traffic signal scheme finally used for timing 1s a discrete integer value, the former having a rounding relationship with the latter. Therefore, when a certain dimension of the particle converges to a certain degree at the local optimal solution, it 1s difficult to jump out of the local optimal position in this dimension.
In addition, most of the existing studies do not use the distribution characteristics of the input traffic flow at the intersection to assist the particle swarm optimization algorithm for the traffic signal optimization. The traditional particle swarm optimization algorithm adopts completely random initialization when initializing the population. However, in the field of traffic signal timing optimization, if we obtain the distribution characteristics of the input traffic flow, we can use this information to purposefully generate an initial population that 1s more likely to cover the area where the optimal solution is located, thereby enhancing the search ability of the algorithm. Therefore, based on the above analysis, the traditional particle swarm optimization algorithm needs to be improved and integrated with the inherent characteristic knowledge in the field of traffic signal timing, so as to enhance the optimization ability of the algorithm in the field of traffic signal timing.
In order to solve the above shortcomings in the prior art, the purpose of the present invention is to provide a method for optimizing smart city traffic signal timing based on multi-exemplar learning particle swarms. This method adopts a multi-exemplar learning particle velocity update strategy, which enhances the diversity of the particle swarm optimization algorithm and helps avoid partict&g03042 falling into the local optimal position. In addition, when the particle population is initialized, the distribution characteristics of the input traffic flow at the intersection are used as the prior-knowledge, so that the initial population distribution is more likely to cover the area where the optimal solution is located.
The purpose of the present invention can be achieved through the follow technical solution:
A method for optimizing smart city traffic signal timing based on multi-exemplar learning particle swarms is provided, comprising the following steps:
S1. The amount and ratio of the traffic flow at each entrance of the intersection is the biggest factor affecting the setting of traffic signals; however, in the existing methods of using intelligent algorithms to optimize the traffic signal timing, the traffic signal timing optimization is regarded as a stochastic optimization problem, and this important factor is not taken into account; in the present invention, an initial seed solution is generated according to distribution characteristics of traffic flow at each entrance of an intersection, and then an initial particle swarm population is generated according to the initial seed solution, the process being as follows:
S101. with the number of lanes at each entrance not necessarily the same, for two entrances with different number of lanes but the same input traffic flow, the entrance with fewer lanes will be more “congested”, and thus theoretically needs relatively more green-signal passage time; therefore, it 1s necessary to define an average lane flow ratio index, which is defined as follows, to measure the traffic flow distribution ratio:
ALE, = 0
LN,
2 ALF LU503042 1° SALE, @ > where ALF; represents the average lane flow at the j-th entrance, JI; represents the input traffic flow at the j-th entrance, LN; represents the number of lanes at the j-th entrance, and R; represents the average lane flow ratio at the j-th entrance;
S102. an initial total green signal duration is calculated according to the ratio of the total input traffic flow of the entire intersection to the maximum traffic flow that the intersection is designed to accommodate, and then the green signal duration is allocated to each entrance to generate the initial seed solution based on the average lane flow ratio, with the green signal duration of the initial seed solution in each dimension calculated by the following formula: g/ =LR,-G'] (3)
ZI,
G' = (4) where 8” represents the green signal duration of the seed solution at the j-th entrance, represents the total green signal duration of the seed solution, G represents the total green signal duration with the maximum allowable traffic signal cycle, Vi. represents the maximum input traffic flow allowed at the intersection, and LJ represents a round down operation; and
S103. the initial seed solution represents the area where the potential optimal timing scheme may exist; based on each dimension value of the initial seed solution, an appropriate Gaussian disturbance is added to generate an initial particle population in the vicinity of the seed solution, with the initial population specifically generated according to the following formula: x/ = g" + N(0,5) (1) where x;represents the j-th dimension of the i-th particle in the population, and
N(0,5) represents a Gaussian distribution with a mean of 0 and a standard deviation of 5; LUS03042
S2. constraint correction is performed on individuals of the generated particle swarm to make them meet the constraint requirements of a feasible solution, which include the maximum and minimum green signal duration 5 constraints, and the maximum and minimum cycle constraints;
S3. after the initial population is generated, the velocity and position of the population can be updated according to the iterative steps of the particle swarm optimization algorithm; however, the traditional particle swarm optimization algorithm has the shortcoming that the particles are prone to fall into the local optimal solution; to overcome this shortcoming, the present invention adds a multi-exemplar learning strategy in the particle velocity update link, and the particles of the population are updated in velocity according to the multi-exemplar learning strategy, with the update formula as follows:
Vv =0v! tor! (pBest! — x!) + cr] (gBest’ —x!) +e,r/ (exemplar’ — x!) (2) where v,…, represents the velocity of the i-th particle in the j-th dimension in the (t+1)-th generation, vw, represents the velocity of the i-th particle in the j-th dimension in the #-th generation, xj, represents a position value of the i-th particle in the j-th dimension in the #-th generation, w represents the velocity weight, pBest; represents the historical optimal position of the i-th particle in the j-th dimension, gBest / represents the global optimal position in the j-th dimension, c1, ¢; and c3 are three update coefficients, r;, 7, and r;are three different random numbers within [0,1] in the j-th dimension, and exemplar represents an exemplar individual value of the i-th particle in the j-th dimension; more specifically, for the exemplar value of the i-th particle in the j-th dimension, two other particles are randomly selected, and the value of the historical optimal position in the j-th dimension of the particle with the better fitness value is selected as the exemplar value of the /-th particle in this dimension; in updating the position value of a particle, the original position plus the updated new velocity in formula (6) is used as the new position of the particle;
S4. the micro-simulation software VISSIM is used to simulate and evaluht@03042 the traffic signal timing scheme represented by the rounded-down position value of a particle, and the evaluation result is used as a fitness value of the particle; and
S5. when the algorithm iterates to a specified maximum number of iterations, the global optimal scheme is used at this time as the final timing scheme, otherwise steps S3 to S4 continue until meeting the maximum number of iterations set in advance by the algorithm, and then the process ends.
Further, in “generating an initial seed solution according to distribution characteristics of traffic flow at each entrance of an intersection” in step SI, when generating the individuals of the population, first taking a random number within [0,1]; if the random number is smaller than K7, generating the individuals according to a knowledge auxiliary strategy, otherwise generating the individuals according to a random method.
Further, in step S3, setting an exemplar update interval threshold RT; if the global optimal position of the algorithm has not been improved after the RT generations, updating the exemplar, otherwise continuing to use the exemplar of the previous generation.
Further, the update coefficients ci, cz and cs are set to 0.75, 0.75 and 1.50, respectively.
Further, the knowledge ratio threshold K7 is set to 0.8.
Further, the exemplar update interval threshold R7'is set to 7.
Further, the maximum number of iterations is set to 40 generations.
Compared with the prior art, the present invention has the following advantages and beneficial effects: 1. The present invention assists the initialization of the particle population by using the distribution characteristics of the traffic flow at the intersection as the prior-knowledge, which helps the initial population to be distributed in the area where the optimal solution 1s more likely to be covered. LUS03042 2. The multi-exemplar learning strategy proposed by the present invention helps particles in different dimensions to learn from different exemplars, respectively, thereby enhancing the diversity of algorithms and preventing particles from falling into the local optimal position during the search process. 3. In the process of generating the initial particle population proposed by the present invention, a knowledge ratio threshold is used to control the ratio of the knowledge auxiliary generation solution to the random solution, which helps to integrate the respective advantages of the knowledge embedding auxiliary strategy and the random search strategy. 4. In the multi-exemplar learning strategy proposed by the present invention, an exemplar update threshold is used to control the update frequency of the exemplar, which, on the one hand, helps to avoid the problem that the particle search is difficult to converge due to the frequent update of the learning exemplars of particles, and on the other hand, helps to timely update the old exemplar individuals that cannot continue to guide the evolution of particles.
Fig. 1 is a flowchart of the multi-exemplar learning particle swarm optimization algorithm according to an example of the present invention; and
Fig. 2 1s a geometric plan view of the intersection of Xinhu Road-Yu’an Ist
Road in Bao’an District, Shenzhen City, China according to an example of the present invention.
In order to make the purposes, technical solutions, and advantages of the examples of the present invention clearer, the technical solutions in the examples of the present invention will be described clearly and completely in combination with the drawings in the examples of the present invention. Obviously, the described examples are some, but not all, examples of the present invention. Ak 03042 other examples obtained by those of ordinary skill in the art according to the examples of the present invention without making any inventive effort shall fall within the scope of protection of the present invention.
As shown in Fig. 1, the method for optimizing smart city traffic signal timing based on multi-exemplar learning particle swarms comprises the following steps:
S1. The traffic signals at the intersection of Xinhu Road-Yu’an 1st Road in
Shenzhen, as shown in Fig. 2, are selected for timing optimization; the intersection is a four-way crossroads and controlled by four-phase traffic signals, with the green signal duration of each phase used as the dimension value of particles; according to the input traffic flow at the four entrances of the intersection, the average lane flow ratio 1s defined as follows:
ALF, A, (3)
LN,
R = A
SALF (4) - where ALF; represents the average lane flow at the j-th entrance, VI; represents the input traffic flow at the j-th entrance, LN; represents the number of lanes at the j-th entrance, and R; represents the average lane flow ratio at the j-th entrance; then, according to the following formulas, the initial seed solution based on the average lane flow ratio is generated, and the green signal duration of the initial seed solution in each dimension is calculated as follows: g’ =LR,-G'] (5)
ZI,
GG (6)
where 8” represents the green signal duration of the seed solution at the PE entrance, represents the total green signal duration of the seed solution, G represents the total green signal duration with the maximum allowable traffic signal cycle, Vi. represents the maximum input traffic flow allowed at the intersection, and LJ represents a round down operation; finally, an initial population is generated according to the following formula based on each dimension value of the initial seed solution: x] =g" + N(0,5) (5) where x;represents the j-th dimension of the i-th particle in the population, and
N(0,5) represents a Gaussian distribution with a mean of 0 and a standard deviation of 5; when initializing the population, a knowledge ratio threshold K7 is set, here set to 0.8 in this example; each time the initial particle is generated by formula (11), a random number in the range of [0,1] is generated first; if the random number is less than K7, the initial particle is generated by formula (11), otherwise the particle 1s initialized randomly;
S2. most of the particles of the initial population in step S1 are generated by
Gaussian sampling based on the seed solution, so it is possible that each dimension value does not meet the requirements of a feasible solution; therefore, it is necessary to perform constraint correction on individuals of the generated particle swarm to make them meet the constraint requirements of the feasible solution;
S3. before the particle 1s updated in velocity, it 1s first determined whether the exemplar individual of the particle needs to be updated; here, an exemplar update threshold RT is set, here set to 7 in this example; if the global optimal position of the algorithm is not improved after the RT generations, the exemplar must be updated before the update of the velocity and position of the particle; the update process of the exemplar is as follows: for the j-th dimension of the exemplar individual, different particles in two other populations are selected, Arr 09042 the j-th dimension of the historical optimal position of the particle with a better fitness value is taken as the j-th dimension of the exemplar individual; if the global optimal position of the algorithm is improved in the R7 generations, it means that the exemplar individual can continue to guide the evolution of particles, so the velocity and position of the particles can be updated directly without updating the exemplar; the particles are updated in velocity according to a multi-exemplar learning strategy, with the update formula as follows: vi, = Ov}, ter! (pBest! — x!) +e,r!(gBest' —x/,)+c,r/ (exemplar! x!) (6) where v,,., represents the velocity of the i-th particle in the j-th dimension in the (t+1)-th generation, v,, represents the velocity of the i-th particle in the j-th dimension in the #-th generation, xj, represents a position value of the i-th particle in the j-th dimension in the #-th generation, w represents the velocity weight, pBest, represents the historical optimal position of the i-th particle in the j-th dimension, gBest / represents the global optimal position in the j-th dimension, exemplar, represents an exemplar individual value of the i-th particle in the j-th dimension, ci, cz and c3 are three update coefficients, and 7, #> and 7; are three different random numbers within [0,1] in the j-th dimension; in updating the position value of a particle, the original position plus the updated new velocity in formula (12) is used as the new position of the particle;
S4. after the particles are updated in the velocity and position, the micro-simulation software VISSIM is used to simulate and evaluate the traffic signal timing scheme represented by the rounded-down position value of a particle, and the evaluation result is used as a fitness value of the particle; and
S5. when the algorithm iterates to a specified maximum number of iterations, which is set to 40 generations in this example, the global optimal solution at this time is rounded down as the final timing scheme, otherwise steps
S3 to S4 continue until meeting the algorithm termination conditions, 1.e. LH 09042 generations as the maximum number of iterations.
To sum up, aiming at the problem that the traditional particle swarm optimization algorithm is prone to fall into the local optimum when applied to the traffic signal timing optimization scene, this method adopts a new multi-exemplar learning strategy, which enables particles to learn from exemplars of other particles in different dimensions while learning from their own optimal position and global optimal position, helping to enhance the algorithm diversity and avoid falling into the local optimal position. In addition, this method adopts the knowledge embedding auxiliary strategy when generating the initial particle population, and uses the distribution characteristics of the input traffic flow at the intersection as the prior-knowledge to assist the generation of the initial population. The optimization results of the traffic flow at a single intersection with different saturation levels show that this method, compared with other timing optimization methods, has better diversity on the premise of ensuring the convergence velocity, and can obtain an optimized timing scheme with better comprehensive performance.
The above examples are preferred embodiments of the present invention, but the embodiments of the present invention are not limited thereto, and any other alterations, modifications, replacements, combinations and simplifications shall be equivalent substitutions and fall within the scope of protection of the present invention.
Claims (7)
- CLAIMS LU503042I. A method for optimizing smart city traffic signal timing based on multi-exemplar learning particle swarm optimization, characterized in that the method comprises the following steps:SI. generating an initial seed solution according to distribution characteristics of traffic flow at each entrance of an intersection, and then generating an initial particle swarm population according to the initial seed solution, the process being as follows: S101. defining an average lane flow ratio as follows: ALE, = 0 LN, R = A SN ALF, (2) > where ALF; represents the average lane flow at the j-th entrance, JI; represents the input traffic flow at the j-th entrance, LN; represents the number of lanes at the j-th entrance, and R; represents the average lane flow ratio at the j-th entrance; S102. according to the following formulas, generating the initial seed solution based on the average lane flow ratio, and calculating the green signal duration of the initial seed solution in each dimension: g’ =LR,-G'] 3) ZI, Gr=6- A ® where 8” represents the green signal duration of the seed solution at the j-th entrance, represents the total green signal duration of the seed solution, G represents the total green signal duration with the maximum allowable traffic signal cycle, Vi. represents the maximum input traffic flow allowed at the intersection, and LJ represents a round down operation; and 17000062 S103. generating an initial population according to the following formula based on each dimension value of the initial seed solution: x/ =g" +N(0,5) (5) where x; represents the j-th dimension of the i-th particle in the population, and N(0,5) represents a Gaussian distribution with a mean of 0 and a standard deviation of 5;S2. performing constraint correction on individuals of the generated particle swarm to make them meet the constraint requirements of a feasible solution:S3. updating the particles of the population in velocity according to a multi-exemplar learning strategy, with the update formula as follows: Vin = @V +0 (pBest/ — x/,)+e,r; (gBest’ — x;,) + e,r; (exemplar! — x) (6) where v’,., represents the velocity of the i-th particle in the j-th dimension in the (t+1)-th generation, vi, represents the velocity of the i-th particle in the j-th dimension in the #-th generation, xj, represents a position value of the i-th particle in the j-th dimension in the #-th generation, w represents the velocity weight, pBest;represents the historical optimal position of the i-th particle in the j-th dimension, gBest / represents the global optimal position in the j-th dimension, exemplar; represents an exemplar individual value of the i-th particle in the j-th dimension, c1, cz and c¢; are three update coefficients, and 7, 7, and r; are three different random numbers within [0,1] in the j-th dimension;S4. using the micro-simulation software VISSIM to simulate and evaluate the traffic signal timing scheme represented by the rounded-down position value of a particle, and using the evaluation result as a fitness value of the particle; andS5. when the algorithm iterates to a specified maximum number of iterations, using the global optimal scheme at this time as the final timing scheme, otherwise continuing steps S3 to S4 until meeting the maximum number of iterations set in advance by the algorithm, and then terminate.
- 2. The method for optimizing smart city traffic signal timing based Lifp03042 multi-exemplar learning particle swarms according to claim 1, characterized in that: in “generating an initial seed solution according to distribution characteristics of traffic flow at each entrance of an intersection” in step SI, setting a knowledge ratio threshold K7, when generating the individuals of the population, first taking a random number within [0,1]; if the random number is smaller than K7, generating the individuals according to a knowledge auxiliary strategy, otherwise generating the individuals according to a random method.
- 3. The method for optimizing smart city traffic signal timing based on multi-exemplar learning particle swarms according to claim 1, characterized in that: in step S3, setting an exemplar update interval threshold RT; if the global optimal position of the algorithm has not been improved after the RT generations, updating the exemplar, otherwise continuing to use the exemplar of the previous generation.
- 4. The method for optimizing smart city traffic signal timing based on multi-exemplar learning particle swarms according to claim 1, characterized in that the update coefficients ci, ¢; and c3 are set to 0.75, 0.75 and 1.50, respectively.
- 5. The method for optimizing smart city traffic signal timing based on multi-exemplar learning particle swarms according to claim 2, characterized in that the knowledge ratio threshold K7 is set to 0.8.
- 6. The method for optimizing smart city traffic signal timing based on multi-exemplar learning particle swarms according to claim 3, characterized in that the exemplar update interval threshold R7 is set to 7.
- 7. The method for optimizing smart city traffic signal timing based on multi-exemplar learning particle swarms according to claim 1, characterized in that the maximum number of iterations is set to 40 generations.
Applications Claiming Priority (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN202110271046.7A CN113012449B (en) | 2021-03-11 | 2021-03-11 | Smart city signal lamp timing optimization method based on multi-sample learning particle swarm |
Publications (1)
Publication Number | Publication Date |
---|---|
LU503042B1 true LU503042B1 (en) | 2023-03-13 |
Family
ID=76406286
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
LU503042A LU503042B1 (en) | 2021-03-11 | 2021-09-06 | Multi-model learning particle swarm-based intelligent city signal light timing optimization method |
Country Status (3)
Country | Link |
---|---|
CN (1) | CN113012449B (en) |
LU (1) | LU503042B1 (en) |
WO (1) | WO2022188387A1 (en) |
Families Citing this family (5)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN113012449B (en) * | 2021-03-11 | 2022-03-29 | 华南理工大学 | Smart city signal lamp timing optimization method based on multi-sample learning particle swarm |
CN116434576B (en) * | 2022-12-12 | 2024-03-12 | 中电信数字城市科技有限公司 | Traffic light timing scheme determining method, device, system and equipment |
CN117523823A (en) * | 2023-10-11 | 2024-02-06 | 吉林师范大学 | Regional traffic signal control optimization method based on quantum genetic algorithm |
CN117077041B (en) * | 2023-10-16 | 2023-12-26 | 社区魔方(湖南)数字科技有限公司 | Intelligent community management method and system based on Internet of things |
CN117787497A (en) * | 2023-12-29 | 2024-03-29 | 西安电子科技大学广州研究院 | Multi-objective optimization method and terminal applied to automobile insurance pricing |
Family Cites Families (6)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN100533475C (en) * | 2006-07-17 | 2009-08-26 | 中国科学院地理科学与资源研究所 | Traffic signal off-line time distribution optimizing method basedon particle group operation method |
CN104134356B (en) * | 2014-06-30 | 2017-02-01 | 南京航空航天大学 | Control method of city intersection model reference self-adaptive signals |
CN106022463B (en) * | 2016-05-13 | 2018-11-13 | 安徽教育网络出版有限公司 | Based on the individualized learning method for optimizing route for improving particle cluster algorithm |
CN106384521A (en) * | 2016-09-18 | 2017-02-08 | 广东工业大学 | Single-intersection traffic signal optimization control method based on public transport priority |
CN109035811B (en) * | 2018-08-28 | 2019-08-20 | 大连理工大学 | A kind of intelligent traffic lamp real-time monitoring method based on digital information element |
CN113012449B (en) * | 2021-03-11 | 2022-03-29 | 华南理工大学 | Smart city signal lamp timing optimization method based on multi-sample learning particle swarm |
-
2021
- 2021-03-11 CN CN202110271046.7A patent/CN113012449B/en not_active Expired - Fee Related
- 2021-09-06 LU LU503042A patent/LU503042B1/en active IP Right Grant
- 2021-09-18 WO PCT/CN2021/119200 patent/WO2022188387A1/en unknown
Also Published As
Publication number | Publication date |
---|---|
CN113012449B (en) | 2022-03-29 |
CN113012449A (en) | 2021-06-22 |
WO2022188387A1 (en) | 2022-09-15 |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
LU503042B1 (en) | Multi-model learning particle swarm-based intelligent city signal light timing optimization method | |
CN112700664B (en) | Traffic signal timing optimization method based on deep reinforcement learning | |
CN108510764B (en) | Multi-intersection self-adaptive phase difference coordination control system and method based on Q learning | |
CN109215355A (en) | A kind of single-point intersection signal timing optimization method based on deeply study | |
CN108538065B (en) | Urban main road coordination control method based on adaptive iterative learning control | |
CN109785619A (en) | Regional traffic signal coordination and optimization control system and its control method | |
CN110136455A (en) | A kind of traffic lights timing method | |
CN104134356B (en) | Control method of city intersection model reference self-adaptive signals | |
CN113538910B (en) | Self-adaptive full-chain urban area network signal control optimization method | |
CN109035767A (en) | A kind of tide lane optimization method considering Traffic Control and Guidance collaboration | |
CN113780624B (en) | Urban road network signal coordination control method based on game equilibrium theory | |
CN111126687A (en) | Single-point off-line optimization system and method for traffic signals | |
CN114995119A (en) | Urban traffic signal cooperative control method based on multi-agent deep reinforcement learning | |
CN114694382B (en) | Dynamic one-way traffic control system based on Internet of vehicles environment | |
Zhou et al. | Fuzzy Control Under Time-Varying Universe and Phase Optimization in Traffic Lights (ICSSE 2020) | |
CN116758767B (en) | Traffic signal lamp control method based on multi-strategy reinforcement learning | |
CN102157065B (en) | Design method of signal intersection provided with straight-through type bus special entrance lane | |
CN110021168B (en) | Grading decision method for realizing real-time intelligent traffic management under Internet of vehicles | |
CN115472023A (en) | Intelligent traffic light control method and device based on deep reinforcement learning | |
CN116564088A (en) | Urban traffic state monitoring and regional signal control system oriented to information physical fusion system | |
CN113096415B (en) | Signal coordination optimization control method for secondary pedestrian crossing intersection | |
CN108597239B (en) | Traffic light control system and method based on Markov decision | |
Wei et al. | Intersection signal control approach based on pso and simulation | |
CN115691110B (en) | Intersection signal period stable timing method based on reinforcement learning and oriented to dynamic traffic flow | |
CN112634619B (en) | Two-type fuzzy single intersection control method and device based on adaptive genetic algorithm and storage medium |
Legal Events
Date | Code | Title | Description |
---|---|---|---|
FG | Patent granted |
Effective date: 20230313 |